A century of wind tunnels since Eiffel

Abstract

Fly higher, faster, preserve the life of test pilots and passengers, many challenges faced by man since the dawn of the twentieth century, with aviation pioneers. Contemporary of the first aerial exploits, wind tunnels, artificially recreating conditions encountered during the flight, have powerfully contributed to the progress of aeronautics.
But the use of wind tunnels is not limited to aviation. The research for better performance, coupled with concern for energy saving, encourages manufacturers of ground vehicles to perform aerodynamic tests. Buildings and bridge structures are also concerned.
This article deals principally with the wind tunnels built at ONERA during the last century. Somme wind tunnels outside ONERA, even outside France, are also evocated when their characteristics do not exist at ONERA.

Full text

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.1 (1-14)
C. R. Mecanique ••• (••••)•••–•••
Contents lists available at ScienceDirect
Comptes Rendus Mecanique
www.sciencedirect.com
A century of fluid mechanics: 1870–1970 / Un siècle de mécanique des fluides : 1870–1970
A century of wind tunnels since Eiffel
Bruno Chanetz
ONERA, Chemin de la Hunière, BP 80100, 91123 Palaiseau cedex, France
a r t i c l e i n f o a b s t r a c t
Article history:
Received 15 November 2016
Accepted 8 March 2017
Available online xxxx
Keywords:
Wind tunnels
Laminar
Turbulent
Subsonic
Supersonic
Hypersonic
Fly higher, faster, preserve the life of test pilots and passengers, many challenges faced by
man since the dawn of the twentieth century, with aviation pioneers. Contemporary of the
first aerial exploits, wind tunnels, artificially recreating conditions encountered during the
flight, have powerfully contributed to the progress of aeronautics.
But the use of wind tunnels is not limited to aviation. The research for better performance,
coupled with concern for energy saving, encourages manufacturers of ground vehicles to
perform aerodynamic tests. Buildings and bridge structures are also concerned.
This article deals principally with the wind tunnels built at ONERA during the last century.
Somme wind tunnels outside ONERA, even outside France, are also evocated when their
characteristics do not exist at ONERA.
2017 Académie des sciences. Published by Elsevier Masson SAS. This is an open access
article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. From birth to first evolutions
1.1. Wind tunnel definition
Wind tunnels are facilities (circular, elliptical or rectangular tunnels) in which the wind is produced by fans or by
compressed air to study and measure the action of the air flow around a solid. The test section is the part of the circuit
where the solid is studied. Invented in the late nineteenth century, these aerodynamic laboratories took off in the early
twentieth century. The method is based on the principle of relativity enunciated by Isaac Newton in 1687: the forces
exerted on a solid immersed in a fluid and the fluid are the same either the solid moves with a certain speed through the
fluid at rest, or the fluid moves, with the same relative velocity to the solid that it is immobile.
In French, wind tunnel is still designated by the term “soufflerie”, which was correct for the first facilities, since a fan was
blowing air upstream (relative to the direction of flow) of the test-section. The first evolution was to suck air downstream
of the test section. The precursors of the other countries in the science of flight use terms that do not prejudge whether the
direction of the air is moving in the circuit: wind tunnel in English, Windkanal in German, galleria aerodinamica in Italian,
and aerodinamicheskaya truba in Russian [1]. Anyway, wind tunnels have largely contributed to the development of aviation,
reducing the number of accidents, thus saving the lives of pilots and maintaining the equipment. They also allowed replacing
the “flair” of the pioneers by the “art” of the engineers, according to the beautiful sentence by Gustave Eiffel.
E-mail address: chanetz@onera.fr.
http://dx.doi.org/10.1016/j.crme.2017.05.012
1631-0721/2017 Académie des sciences. Published by Elsevier Masson SAS. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.2 (1-14)
2B. Chanetz / C. R. Mecanique ••• (••••)•••–•••
Fig. 1. Aérodrome of Captain Ferdinand Ferber in Chalais–Meudon (© ONERA).
1.2. Competitors for ground investigation
The wind tunnel is a very convenient experimental investigation mean, which has surpassed the alternative methods
based on the direct movement of the solid in the air. However, at the beginning of the twentieth century, there was no
consensus. Indeed Armand de Gramont, Duke of Guiche, who was measuring the distribution of pressure on a wing fixed on
a moving car, was a detractor of Eiffel, who was using a wind tunnel. He even challenges the work of Eiffel and disapproves
the transposition of the results obtained in his wind tunnel to the reality of airplanes in flight. To definitively close the
controversy about the equivalence between real and relative movements – although the question had been already solved
in the middle of the 19th century through the experiences of Duchemin, solving the paradox of du Buat [1] – Eiffel requests
the intervention of the great mathematician Henri Poincaré. Shortly before the death of Poincaré in 1912, Eiffel received a
note in line with his expectations: “There is no reason that the forces exerted on the plate by a uniform air flow differ from
those that would occur if this plate is moving in a calm air”. Poincaré adds: “it is clear that only the relative movement is
significant” [2].
Below are described these alternative processes existing at the beginning of the 20th century:
– Horizontal rectilinear motion
•in 1901, a flight testing was performed by the German company Siemens on a train traveling at 160 km/h,
•in 1909, thanks to a 1.4-km private railway track, the AeroTechnical engineering Institute (IAT) of Saint-Cyr-l’École tested
wings;
•during the first decade of the twentieth century, Armand de Gramont, duke of Guiche, tested wings in his car [2].
– Vertical rectilinear motion
•in 1908, before the building of its first wind tunnel, Gustave Eiffel conceived a very ingenious drop test machine to
study the drag of solids [2,3]. He installed the device on the eponymous tower’s second floor, taking advantage of
its 115 m in height. His first studies, awarded by the French Academy of Sciences in 1908, allowed him to found
the fundamental laws of air resistance. In this period, with the growth of aviation, Eiffel brought new ideas and his
knowledge of aerodynamics led him to understand the efforts of the air on a solid.
– Combination of the horizontal and vertical motions
•In 1904, Ferdinand Ferber in Meudon performed a device by taking advantage of the effects of gravity to move a plane
along a cable. His plane, suspended from a sliding carriage along a tensioned cable between pylons (see Fig. 1). Gustave
Eiffel had also considered such a device, called aérodrome, from the first floor of the tower, prior to the realization of
its wind tunnel.
– Circular motion:
In 1906, the AeroTechnical engineering Institute (IAT) of Saint-Cyr-l’École tested solids attached to the ends of rotary
arms. This way allows access to tangential speeds all the more important that the arm is long and the speed great.
1.3. The blowing wind tunnels and the introduction of the collector
Even if the first wind tunnel known in Great Britain operated by compressed air ejector (Venham Francis, 1871, then that
Horatio Phillips, 1884), the following wind tunnels used the air moved by a fan disposed upstream of the wind tunnel’s
test section (see Fig. 2a): Charles Renard (France, 1896), Hiram Maxim (Great Britain, 1896), Konstantin Tsiolkovsky (Russia,

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.3 (1-14)
B. Chanetz / C. R. Mecanique ••• (••••)•••–••• 3
Fig. 2. a: Fan upstream of the model. b: Fan downstream of the model.
1897), Orville and Wilbur Wright (USA, 1901), and Auguste Rateau (France, 1909). We should also mention the wind tunnel,
smoke machine built by Étienne-Jules Marey (France, 1899), which permitted the realization of the first visualizations of
fluid flows [4].
A first development consists in the introduction of the collector. Disposed upstream of the test section, the convergent
part increases the velocity in the test section under the mass conservation law for the flow, which is expressed for an
incompressible flow, by:
V·S=constant
where Vrepresents the velocity of the flow and Sthe surface in the considered section.
1.4. The introduction of suction
The passage to suction is also an important step in the art of wind tunnels. The fan is now disposed downstream of the
test section (see Fig. 2b) and thereby no longer interferes with the model, which is a great improvement to guarantee the
quality of flow. According to this principle the following wind tunnels were conceived: Nikolai Zhukovsky (Russia, 1902),
Thomas Stanton (Britain, 1903), Dimitri Riabouchinsky (Russia, 1905), Gustave Eiffel (France, 1909). Suction wind tunnels
have been developed very quickly. There are currently several hundred major facilities of this kind in the world, while only
less than a dozen were active before 1914.
1.5. Eiffel’s wind tunnels and the introduction of the diffuser
In 1909, Eiffel built its first wind tunnel at the foot of the Eiffel Tower in the Champ de Mars. This facility was operational
till 1911 to study aerodynamics. The air was moved by a fan driven thanks to a 50-horsepower electric motor. The first
tests were to check that the principle of wind tunnel simulation – where the aircraft or its model lying in air in motion –
is consistent with the reality – where the aircraft is moving in the air considered as non-moving as a first approximation.
To this aim, Eiffel compares the wind tunnel results with those previously obtained with the drop test machine. This
comparison being successful, Eiffel’s wind tunnel was then available for the pioneers in their conquest of air: Farman,
Bleriot, Voisin, Bréguet [5].
In 1911, the municipality of Paris wanting to recover the area on which the wind tunnel is located, Eiffel had to leave
the Champ de Mars. In the beginning of 1912, he installed in the district of Auteuil a new wind tunnel with increased
performance, this new laboratory being inaugurated on 19 March. The originality of the wind tunnel at Auteuil is the
presence of the following successive elements, from upstream to downstream:
– a collector or convergent, where air is accelerated from an entrance section, four meters in diameter, to an outlet of
two meters;
– a closed experimental chamber where the free stream flow is produced (i.e. without walls); whereas at the time the
other wind tunnels had a guided test section;
– a recovery at the exit of the test chamber, made of a small collector whose section is larger than the section of the
convergent at the test chamber entrance;
– a diffuser made of diverging tube in order to recover pressure in front of the fan (see Fig. 3). The same 50-horsepower
engine as that already used in the Champ de Mars, was used again. The wind tunnel of the Champ de Mars reached a
speed of 18 m/s in a 1.5-m-in-diameter test section. With the same engine, the new wind tunnel, thanks to the diffuser,
reached 32 m/s (an increase of 78% speed) in a section of 2 meters in diameter (an increase of 78% also), hence a mass
flow gain of more than 200%.
– A helical fan with a diameter of 3.80 m and a weight of 8.500 kg was specially purchased for the wind tunnel of Auteuil,
the fan of the Champ de Mars wind tunnel being used for a second small wind with a collector outlet one meter in
diameter.

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.4 (1-14)
4B. Chanetz / C. R. Mecanique ••• (••••)•••–•••
Fig. 3. Introduction of the diffuser between the test section and the fan.
Fig. 4. Watercolor drawing of the Eiffel wind tunnel located in the rue Boileau (© Aérodynamique Eiffel).
The diffuser, which was an innovation of Eiffel (see Fig. 4), was the subject of a patent dated 28 November 28 1911:
“The addition of a diffuser to improve the performance of a machine to produce artificial air flow”. This invention was rich
in consequences, as it allowed Eiffel to drastically reduce the electric power required for such a facility. The efficiency of
the device comes from the Bernoulli law, stating that pressure and velocity vary inversely. Indeed, by reducing velocity, the
diffuser increased the pressure of the air exiting the test section. The pressure difference on either side of the fan is then
much less than the one that would exist if the fan were located directly just downstream of the test-section. This device
leads to a reduction in the electric power required to extract the air. “Thus the diffuser saves money since power is reduced
by a factor of 3. The advantage of this recovery device is clear, which allowed us to realize the present large facility” (Eiffel
[5]). From that date all wind tunnels have been equipped with a diffuser. Gustave Eiffel (see Eiffel in Fig. 5) will ask that
the official designation, “Aerodynamic device system Eiffel Paris” should be engraved on a 1-m-by-0.5-m marble plaque on
all facilities using his technique, without asking for royalties.
1.6. The Prandtl wind tunnel in Göttingen
Another major innovation consisted in making air circulate in a closed circuit. The first installation of this kind was
built at the University of Göttingen, in Germany, by Ludwig Prandtl in 1909. After being sucked downstream of the test
section, the air is guided through four successive bends, and then readmitted upstream of the collector. This type of wind
tunnel, called Prandtl wind tunnel – or Göttingen wind tunnel – leads to a better fuel efficiency and provides controlled
test conditions (pressure, temperature, humidity). Modern wind tunnels combine Prandtl’s innovation (closed circuit) and
Eiffel’s one (diffuser).
1.7. The controversy between Eiffel and Prandtl concerning the drag of spheres and the irruption of the Reynolds number in
aerodynamics
At the Auteuil laboratory, Eiffel continued the activities undertaken at the Champ de Mars Laboratory concerning the
aerodynamics of the drag of a sphere. The measurements carried out at the Champ de Mars’ wind tunnel at a speed of
15 m/s revealed values of the drag coefficient less than half those found for lower speeds by Professor August Föppl in
the Göttingen Laboratory, headed by the famous Ludwig Prandtl. Föppl did not hesitate to write that the French colleague
had committed an error, but a forgivable mistake with respect to its age. Therefore, Eiffel decided to undertake new tests

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.5 (1-14)
B. Chanetz / C. R. Mecanique ••• (••••)•••–••• 5
Fig. 5. Eiffel in the vacuum chamber of its wind tunnel in the Boileau street (© Aérodynamique Eiffel).
with spheres of different diameters. He discovered that, for each sphere, there are two flow regimes: one at low speeds
corresponding to the coefficient found in Göttingen (with boundary layer separation in laminar flow) and the other, at higher
speeds, corresponding to the coefficient found in the Champs de Mars facility (with boundary layer separation in turbulent
flow). The transition between the two regimes always occurs for the same value of the product velocity ×diameter.
If the viscosity is constant, which is the case at low velocities, this product velocity ×diameter is similar to the Reynolds
number:
Re =ρ·V·L/µ
where ρis the density of the actual flow, Vits velocity, µthe air dynamic viscosity, and Lthe characteristic length of the
studied solid.
Thus, the experimental studies carried out in his laboratory of Auteuil have first demonstrated the important role of the
Reynolds number in aerodynamics.
The example of the sphere reveals the complexity of the boundary-layer behavior, which sometimes leads to paradoxical
laws. The laminar boundary layer generates a lower friction drag than the turbulent one, but its earlier separation leads
to a more disturbed wake, which is the origin of a form (pressure) drag larger than in the turbulent case. Therefore, by
triggering transition, one reduces more the drag due to the wake than one increases the friction drag due to a more
important turbulent boundary layer on the sphere. The turbulent regime is more favorable to the global drag balance, which
is the sum of the friction and form drags. This phenomenon is illustrated by an old experiment carried out at ONERA in
a water tunnel by Werlé [6]. Fig. 6a is relative to the case obtained for a Reynolds number based on the diameter equal
to 200,000, and Fig. 6b is relative to the case of a Reynolds number equal to 300,000. One sees clearly the less perturbing
wake in the second case (with a turbulent separation) than in the first case (with an earlier laminar separation) [7,8].
An emblematic application of the phenomenon is given by the golf ball with dimples that trigger transition and postpone
more downstream the separation of the boundary layer. The reduced wake resulting from this turbulent boundary layer
promotes the penetration in the air by reduction of the global drag, which leads to an increase in the golf ball range. At
the origin, the inventors of golf manufactured leather balls, taking care to place the seams on the back to have a smooth
envelope favorable to penetration into the air. They later discovered that the best golf balls were damaged balls whose
rough envelope triggers a transition of the boundary layer [8,9].
The controversy between Prandtl and Eiffel about the drag of the sphere highlights the essential role played by exchanges
between researchers for the advancement of knowledge. Thanks to these confrontations, science progresses and that was the
goal pursued by Gustave Eiffel: “It is mainly the progress in aviation that I wanted by creating this laboratory where the tests

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.6 (1-14)
6B. Chanetz / C. R. Mecanique ••• (••••)•••–•••
Fig. 6. a: Laminar regime at the boundary layer separation. b: Turbulent regime at the boundary layer separation (© ONERA).
are absolutely free, but where the results would be published, in the general interest, through reports or communications
to scientific societies”. Gustave Eiffel was a great researcher who has not played the role of a suspicious inventor against
potential competitors. He offered all manufacturers the means to test their models free of charge, if their results were to
be published and accessible to all. It was truly the wind of research that was blowing at the laboratory of Auteuil under
the leadership of Eiffel. He publishes The Friction of the Air and Aviation, a book translated into both German and English.
Furthermore, its contribution to this emerging science of aerodynamics was recognized in the USA, where he received the
gold medal of Langley in 1913, which had only been awarded to Wilbur and Orville Wright before.
2. Facilities from the beginning of the 20th century to modern wind tunnels
2.1. The rules of similarity
During the twentieth century, the development of wind tunnels has continued, driven by the urgent need to respect the
similarity rules to ensure the validity of the tests. A first evident requirement was to have a model with the same geometric
shape as the original object.
A second condition was required to meet the characteristics of the flow: the Mach number Mhas to be respected:
M=V/c
where Vrepresents the velocity of the flow and cthe celerity of sound.
The third similarity rule is to perform tests maintaining – as far as it might be – the Reynolds number (Re) of the flow.
If the air in the wind tunnel has the same characteristics as the real one (density and viscosity) and the same velocity,
the Reynolds number is automatically respected in a model at the real scale. Since at the beginning of the epic of aviation,
the size of aircrafts was small, the first mean to reach the convenient Reynolds number being to increase the size of the
wind tunnels. The large Chalais–Meudon S1Ch wind tunnel meets these specifications.
Due to the rapid increase in aircraft size, it was not suitable to increase too much the size of the facilities. Thus, nowadays
the experiments are again performed with reduced models. To respect the convenient Reynolds number, we play on the
other parameters of the formula: density by pressurizing the fluid and viscosity by cooling the fluid. The F1 wind tunnel is
a pressurized wind facility and the European Wind Tunnel (ETW) is a cryogenic wind tunnel.
2.2. The large Chalais–Meudon S1Ch wind tunnel (France)
Antonin Lapresle, collaborator and successor of Eiffel at the Auteuil wind tunnel, designed, at the request of the Air
Ministry, the large Chalais–Meudon wind tunnel (Hauts-de-Seine) that had been built between 1932 and 1934 for testing
full-scale airplanes (see Figs. 7 and 8). It was equipped with a collector with a reduction ratio of 3.5, with a mouth of
350 m2that captures the outside air and allows it to reach a velocity of 180 km/h in the inlet test section of 100 m2. The
diffuser, a giant tube shaped like a truncated cone (elliptical section, 10 m of vertical axis and horizontal axis of 18 m), is
made of reinforced concrete with a thickness of 70 mm and a length of 38 meters. Downstream of the diffuser, the suction
chamber, with six fans of 1,000 horsepower each, allows the air extraction. The installation, which had become obsolete, has
been disused in 1977. Classified on the list of the historical monuments in 2000, this facility has tested, in many national
programs, aircraft, automobiles and buildings aerodynamics [10].
2.3. The F1 wind tunnel of Le Fauga–Mauzac (France)
To replace the large S1Ch wind tunnel of Chalais–Meudon, other facilities were emerging. For aerospace testing, a pres-
surized F1 wind tunnel was built near Toulouse, at Le Fauga-Mauzac. Operational since 1974, this tunnel closed circuit has

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.7 (1-14)
B. Chanetz / C. R. Mecanique ••• (••••)•••–••• 7
Fig. 7. View of the large S1Ch wind tunnel in Meudon (© ONERA).
Fig. 8. Bréguet 940 airplane in the S1Ch Meudon test section (© ONERA).
a rectangular test section (4.5 m ×3.5 m) where the flow speed reaches 430 km/h. Pressurizing the facility (up to 3.85 bar)
increases the density of the flow, hence maintaining a high Reynolds number compensating the reduced size of the model
imposed by the relatively small wind tunnel size. It has not been, however, possible to save more than a factor of 4 on the
Reynolds number by pressurizing, since then very large dynamic loads will appear on the structures (walls, floors).
2.4. The European Transonic Wind Tunnel (ETW) in Cologne (Germany)
To obtain higher number of Reynolds, close to the one that characterizes the current large civil aircraft, we chose to play
on the air’s dynamic viscosity µ, which is proportional to the square root of temperature. By cooling the circuit, a cryogenic
wind tunnel is obtained. In such a facility, the Reynolds number increases not only due to the decrease in viscosity, since
it is included in the denominator in the formula giving the Reynolds number, but cooling also induces an increase in the
density via the ideal gas law:
ρ=p/r·T
where ρis the density in kg m−3, pthe pressure in Pa, rthe universal gas constant equal to 287 J kg−1K−1and Tthe
temperature expressed in K.

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.8 (1-14)
8B. Chanetz / C. R. Mecanique ••• (••••)•••–•••
Fig. 9. View of the Mach 2 nozzle of the S5Ch wind tunnel in Meudon (© ONERA).
The decrease in temperature contributes also to an increase in the Reynolds number, via the increase of density. Such
a system was installed in Cologne, Germany. This is the ETW wind tunnel in which the flow is at a very low temperature
obtained by cooling the air by means of an internal circulation of liquid nitrogen at 100 K between the walls. This facility
reached Mach 1.35. This is a transonic wind (flow velocity around Mach 1) put into operation in 1994.
3. Supersonic and hypersonic wind tunnels
3.1. Special features of supersonic flows
The race for speed has required the construction of more efficient facilities. This situation led to the creation of transonic
wind tunnels (around Mach 1), supersonic (Mach 1 to Mach 4), and hypersonic (from Mach 5). However, supersonic flows
obey laws fundamentally different from those governing subsonic flows. The Hugoniot relation distinguishes these two
states. It expresses, in the absence of friction, the relationship between the Mach number Mof the flow, the velocity
variation dVof the fluid along the axis of the wind tunnel and the section variation dA:
dA
A=³M2−1´dV
V
Thus, in subsonic flow (M<1), the factor (M2−1) is negative, which leads to opposite variations of Aand V. Thus, the
speed increases in the converging sections (collector), since the section of the tunnel decreases, and decreases in diverging
sections as in the subsonic diffuser described above.
However, for the supersonic flow (M>1), the factor (M2−1)becomes positive. It follows that the velocity increases in
the diverging sections (since the section of the tunnel increases) and decreases in the converging sections (since the tunnel
section decreases).
Another consequence of this relationship consists in obtaining the sonic flow (M=1)in a minimum area section called
throat. The air is moved by a fan or other means to suck the air in the circuit.
3.2. Supersonic wind tunnel
Due to the particularity of the flows, the realization of a supersonic wind tunnel requires the successive establishment
of a convergent device that increases the velocity of flow to reach Mach 1 at the throat, itself followed by a diverging
portion (called nozzle) within which the speed becomes supersonic. This convergent-divergent device was first designed by
the Swedish engineer Carl Gustav de Laval in 1890 [8]. A Swiss engineer, Jacob Ackeret, built from this concept the first
supersonic wind tunnel in 1935 [11]. One can see in Fig. 9 a view of the Mach 2 nozzle in the S5Ch wind tunnel in ONERA
Meudon, into operation since 1954.
The supersonic flow has its share of mystery and it was believed for a long time that we could not reach the speed of
sound, hence the usual expression “wall of sound”. This is due to the drastic increase in drag that opposes the advance-
ment near Mach 1. Also, the increase in propulsive power planes was balanced by a corresponding increase in drag, which
suggested that this barrier was impassable.

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.9 (1-14)
B. Chanetz / C. R. Mecanique ••• (••••)•••–••• 9
Fig. 10. The large ONERA S1MA wind tunnel in Modane (© ONERA).
3.3. Hypersonic wind tunnel
Obtaining flows higher than Mach 4 requires a preheating of the air prior to the expansion in the nozzle in order to avoid
air liquefaction. Thus flow streams from Mach 4 or Mach 5 are often called hypersonic flows, although specific hypersonic
features, such as the dissociation of molecules, appear at higher Mach numbers [12].
The hypersonic domain mainly concerns the field of rockets or space vehicles. Besides the need to heat the air upstream
of the nozzle, hypersonic wind tunnels also require more power to “start” the throat, i.e. to obtain Mach 1 at the throat.
Indeed, it is necessary to establish a pressure difference between the upstream and downstream ends of the nozzle, suffi-
cient to create the conditions for the ruin. Thus it is required to obtain high compression upstream or intense depression
downstream, or the combination of both effects. In fact, most of the time, for hypersonic applications, one builds blow-down
wind tunnels. The wind tunnels for lower Mach number are usually continuous wind tunnels, able to work during several
hours. At the opposite, a blow-down wind tunnel operates for a short time (from a few seconds to a few minutes), which is
determined by the capacity of the compressed air reservoirs upstream and/or the capacity of the vacuum tanks downstream.
4. Wind tunnels for aircraft
4.1. The S1MA wind tunnel in Modane–Avrieux (France)
Among the hundreds of wind tunnels in service worldwide, the most impressive is certainly the S1MA wind tunnel at
ONERA in Modane–Avrieux in Savoy, France [13–15], with a power of 88 megawatts (MW), higher than that of the Charles
de Gaulle aircraft carrier. His performances are due to the German engineers who had started its construction in the Austrian
Alps during Second World War. Unfinished at the end of the latter and considered as a compensation for war damage, it
was transferred to France and was put into service in 1952 (see Fig. 10).
The air flow is provided by two wheels with a diameter of 15 meters (see Fig. 11) rotating in opposite directions and
provided with metal blades (10 for one and 12 for the other). Each of these two fans is driven by a 44-MW Pelton turbine.
To supply water to the turbines, 10 million m3of water per year come from the mountains thanks to several barrages
operated by EDF, the French power company. Since the barrage overlooking Avrieux, a 840-meter waterfall feeds the wind
tunnel through a conduct, the flow rate reaching 15 m3/s. In this closed circuit wind tunnel, the speed can reach 1100 km/h
(Mach 1) in the test section having a diameter of 8 meters, giving it the status of the largest sonic wind tunnel reaching the
speed of sound in the world.
4.2. The icing wind tunnel in Capua (Italy)
Thanks to the climate conditions prevailing in Modane due to the altitude, it was possible to study the icing phenomena
that affect airplanes as they pass through the clouds. However, the S1MA wind tunnel at Modane is no longer used for

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.10 (1-14)
10 B. Chanetz / C. R. Mecanique ••• (••••)•••–•••
Fig. 11. The fan in the S1MA wind tunnel in Modane (© ONERA).
this type of test, because it could not guarantee permanent – even in winter – temperature conditions to obtain this
phenomenon.
Thus the experimental study of icing now requires, for any time of the year, the use of wind tunnels specially equipped
to produce airflow with water droplets. A very important test facility dedicated to this type of study is that of the Italian
Aerospace Research Centre (CIRA) in Capua, near Naples. This huge facility, into operation since 1985, occupies a floor area
of approximately 2,500 m2, and the test section is 11.40 m long. It is equipped with a water spray system that generates
various types of clouds, including reproducing the reality of flight for a civil aircraft flying at an altitude of 10 km with an
outside temperature of −50 ◦C. Under these conditions, the condensation of water droplets contained in the clouds can form
a layer of ice on various elements of the surface planes if precautions are not taken to prevent this phenomenon. However,
it is important to maintain the original shape of the wing, optimized to ensure the aircraft’s flying qualities. Therefore it is
imperative to study in wind tunnels local devices that protect sensitive areas against the formation of frost.
5. Wind tunnels for spacecraft
5.1. The S4MA wind tunnel in Modane (France)
High velocity wind tunnels are required for the specific needs of strategic missiles, rockets and space vehicles. During a
re-entry, velocities encountered are in the order of 20,000 km/h. The first hypersonic wind tunnel category consists of wind
tunnels called “cold”, i.e. the air is heated before its expansion in the nozzle, just enough to prevent liquefaction. Thereby,
the air – or any other study gas – reaches few tens of kelvins (around −250 ◦C). The S4MA wind tunnel operates on this
principle since 1970. It is constituted, from upstream to downstream by:
•a compressed air tank at 270 bar,
•a Joule effect heater amount up to 1600 ◦C,
•a set of three interchangeable nozzles: Mach 6.4, Mach 10 and Mach 12,
•a test chamber,
•a diffuser for compressing the air,
•a vacuum sphere for sucking the air ejected from the diffuser.
This tunnel allows runs with a duration time reaching 45 s at Mach 12. In such a “cold” wind tunnel, the simulation of
real phenomena of hypersonic flight is not fully ensured, since the velocity achieved is still insufficient, although reaching
6500 km/h with a Mach 12 nozzle. Furthermore, in the reality, due to the air friction on the vehicle, which generates
an intense temperature increase, the dissociation of oxygen and nitrogen molecules occurs. In a cold wind tunnel, the air
temperature is too low and the air still retains its molecular structure, and is not the seat of the reactive phenomena actually
observed.
The shape of the nozzle permits to obtain the desired Mach number, but the temperature conferred to the flow deter-
mines the velocity (V) at the nozzle outlet. This fact is due to the definition of the number of Mach (M), since the velocity
V=M·c, with cthe sound speed, which is itself proportional to the square root of the temperature.
The dependence of the Mach number on the temperature can be conveniently illustrated by the example of Concorde
at its cruising velocity. He was flying at Mach 2 at 18 km of altitude, where the temperature is around −50 ◦C. Thus the
aircraft velocity at that distance from the ground reached 2110 km/h. If the plane had kept the same velocity at sea level,
with an outside temperature of +20 ◦C, this would lead to a supersonic flow only reaching Mach 1.7 instead of Mach 2.

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.11 (1-14)
B. Chanetz / C. R. Mecanique ••• (••••)•••–••• 11
Fig. 12. The high-enthalpy ONERA F4 wind tunnel in Le Fauga–Mauzac (© ONERA).
In fact, with a “cold” wind tunnel, one easily gets high Mach numbers, since the sound speed is low, approximately
350 km/h instead of 1200 km/h in standard conditions for temperature and pressure. Thus the resulting Mach number may
be high and the gas velocity still important, but not entirely representative of the reality of an atmospheric reentry.
5.2. The F4 wind tunnel in Le Fauga–Mauzac (France)
To reproduce faithfully the reality of hypersonic flights, one uses “hot” hypersonic wind tunnels, called hot-shot wind
tunnel, in which the test gas is heated intensely upstream of the nozzle. But considerable powers are required to obtain
the initial temperatures, i.e. stagnation temperatures, before expansion in the nozzle exceeding 2500 ◦C, with high pressure
levels. So the run duration is extremely short. In the F4 wind tunnel, into operation since 1992 (see Fig. 12), the heating of
the flow is obtained by an electric arc heater in a 15-liter chamber, filled with air, nitrogen or other test gases at a selected
pressure depending on the desired final conditions. The burst of a diaphragm located at the throat allows the admission
of the gas flow in the nozzle (up to Mach 20). Thus one produces pressures of several thousand bars and temperatures of
3000 to 8000 ◦C, for useful run duration in the order of magnitude of 100 ms. The energy required for the realization of
the electric arc is obtained by means of a generator consisting of a 15-ton flywheel launched at 600 revolutions per minute,
which represents an energy storage equal to 400 MJ. Since these facilities require significant energy, the test sections are of
small dimensions, with a diameter of a few tens of centimeters (according to the nozzle used).
6. Wind tunnels for ground vehicles
6.1. Cars in wind tunnel: a long-standing idea
Testing cars in a wind tunnel is not a new idea. In 1914, Peugeot passed a car in the Eiffel wind tunnel located in the rue
Boileau, in Paris. The left view (Fig. 13a) is relative to the first model tested and the right view (see Fig. 13b) is relative to
the model improved after wind tunnel tests. To decrease the aerodynamic drag, the back shape of the car has been largely
modified.
With the requirements for reducing pollutant emissions, these tests are more and more relevant. Indeed the aerodynamic
drag becomes important from a 90 km/h cruise velocity. A gain of 10% in the aerodynamic drag leads to a decrease in
consumption of 0.3 to 0.4 liters per 100 km on the highway at 130 km/h.
6.2. The S2A wind tunnel in Montigny-le-Bretonneux (France)
In 2003, to meet the needs of the automobile industry, a highly advanced wind tunnel, with a capacity of 3.8 MW, was
put into service in Montigny-le-Bretonneux (Yvelines, France) by an economic interest grouping involving Peugeot, Renault,
and the “Conservatoire national des arts et métiers” (CNAM). This wind tunnel and aeroacoustics (S2A) reaches a velocity of
240 km/h in a test section of 24 m2, allowing one to study aerodynamics and acoustics on vehicles at scale 1 (see Fig. 14).

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.12 (1-14)
12 B. Chanetz / C. R. Mecanique ••• (••••)•••–•••
Fig. 13. Peugeot model tested in 1914 in the Eiffel wind tunnel (© Laboratoire Eiffel). a) Before wind tunnel tests. b) After wind tunnel tests.
Fig. 14. S2A wind tunnel in Montigny-le-Bretonneux (© GIE S2A).
For the simulation of the ground effects, this wind tunnel is equipped with a central moving carpet under the vehicle, which
rolls at the same wind tunnel velocity.
7. Wind tunnels for buildings
7.1. The Eiffel wind tunnel in Paris (still in service)
Stadiums, bridges or towers, all these works have in common their gigantism. Wind effects on these structures are the
subject of extensive studies. Gustave Eiffel wind tunnel ever works, principally in this domain, since the French Scientific
and Technical Centre for Building (CSTB) bought the facility in 2001. Also added to the wind resistance for buildings, a lot
of other studies have been undertaken in the facility, such as effluent dispersion in nature, active cooling (with pumps to
move air), passive cooling (without pumps), which are subject of extensive research in the current context.
7.2. The Jules Verne wind tunnel in Nantes (France)
CSTB operates also, since the end of the last century an exceptional large facility: the Jules Verne climatic wind tunnel
in Nantes (see Fig. 15). This tunnel was conceived to study the combined effects of wind and other climatic parameters
(rain, sand, sun, snow, temperature...) on building elements, but also on vehicles or other system submitted to extreme
climatic conditions. This installation consists of two concentric rings. A dynamic circuit occupies the outer ring, where six
fans deploy a 3-MW total energetic capacity, a velocity of 100 km/h reached in the 100-m2test-section of 100 m2and
280 km/h in the “high-velocity” 30-m2test section. The wind tunnel can simulate heavy rain (rainfall up to 200 mm/h) and
sandstorm. The thermal circuitry occupies the inner ring and simulates the temperature (−2◦C to +50 ◦C), fog, ice, snow,
sunshine and all types of rain. The velocity reaches 140 km/h using a fan with a 1.1-MW electric power. Heat exchangers
with a cooling capacity of 2 MW ensure the control and the variation of the temperature till 15 ◦C/h. The models of the
bridges of Normandy and Millau have been tested in wind tunnels.

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.13 (1-14)
B. Chanetz / C. R. Mecanique ••• (••••)•••–••• 13
Fig. 15. The “Jules Verne” wind tunnel in Nantes (© CSTB).
8. Conclusion
In the 1980s, the advent of computers with large computing capabilities, in conjunction with the development of nu-
merical methods, augured the disappearance of wind tunnels. However, the abandonment of these aerodynamic ground
laboratories is still not in the agenda. Yet thanks to the evolution of computing, Computational Fluid Dynamics (CFD) has
now reached a high degree of confidence, so that the researcher in aerodynamics considers that it is an excellent good
means to understand the physical reality as well as the measurements acquired during flight tests or in wind tunnels.
Parallel to the development of theoretical models, another breakthrough, almost as spectacular, took place in the exper-
imental area with the development, thanks to laser sources, of a set of optical means to explore complex flows [16–18].
With this breakthrough of metrological means, wind tunnels allow one today to assess the validity of the theoretical models
proposed in fine physical flows. Moreover, taking advantage of the constantly improved reliability of digital means, the con-
cept of computer-aided wind tunnels was born. This process aims to correct the results of tests by the results of numerical
calculations in conjunction with the tests in the same wind tunnel environment. The characteristics of the test are taken
into account in order to determine the effect of the walls, the effect of the support and, more generally, of anything that
could affect the experiment. Hence, calculation and experiment will still bring a mutual support in the years to come.
Furthermore, the numerous wind tunnel construction projects in emerging countries show that the use of experiment in
the aerodynamics domain is still relevant. This article has been principally devoted to ONERA wind tunnels. To have more
information about aerodynamic research in France, see [19] and in foreign countries, the reader is referred to [20].
References
[1] B. Chanetz, Les souffleries, La science au présent, Encyclopaedia Universalis, 2015.
[2] M. Peter, J.-P. Cuisinier, Eiffel: la bataille du vent, CSTB, 2007.
[3] C. Fontanon, Eiffel, Darrieus et l’aviation, in: J.-F. Belhoste (Ed.), Le Paris des Centraliens, Direction de l’action artistique de la Ville de Paris, Paris, 2004,
pp. 185–188.
[4] M.-L. Théodule, L. Mannoni, B. Chanetz, Marey, précurseur oublié des souffleries, La Recherche 380 (November 2004) 66–71.
[5] G. Eiffel, Nouvelles recherches sur la résistance de l’air et l’aviation faites au laboratoire d’Auteuil, 1914, H. Dunot et E. Pinat, éditeurs.
[6] H. Werlé, Transition et turbulence, Note technique Onera 1987-7, FR, ISSN 0078-3781.
[7] B. Chanetz, M. Peter, Gustave Eiffel, a pioneer of aerodynamics, Int. J. Eng. Syst. Model. Simul. 5 (1–3) (2013).
[8] J.-C. Lengrand, B. Chanetz, Boundary Layer Lecture, École centrale de Paris, third study year, M.A.E option Air-Space, 2005.
[9] M.-C. Coët, B. Chanetz, M. Peter, Eiffel, pionnier de l’aérodynamique, Centraliens 617 (April 2012).
[10] M. Bazin, J. Carpentier, B. Chanetz, M.-C. Coët, J. Délery, J.-P. Marec, De l’aérostation à l’aérospatiale, Centre de recherche de l’Onera, ONERA, Meudon,
2007.
[11] B. Chanetz, A. Chpoun, Supersonic and hypersonic wind tunnels, in: G. Ben-Dor, O. Igra, T. Elperin (Eds.), Handbook of Shock Waves, Theoretical,
Experimental and Numerical Techniques, vol. 1, Academic Press, 2001, pp. 651–682, Chapter 4.5.
[12] V. Lago, A. Chpoun, B. Chanetz, Shock waves in hypersonic rarefied flows, in: R. Brun (Ed.), High Temperature Phenomena in Shock Waves, Springer,
2012, Chapter 7.
[13] B. Chanetz, M.-C. Coët, J. Tensi, La grande soufflerie de Modane, in: PEGASE, la revue du musée de l’Air, No. 137, June 2010.
[14] M. Pierre, Création du centre d’essais de l’Onera à Modane–Avrieux, Onera, No. 777, April 1987.
[15] M. Pierre, Développement du centre d’essais de l’Onera à Modane-Avrieux, Onera, No. 1015, October 1995.
[16] J. Délery, B. Chanetz, Experimental aspects of code verification/validation: application to internal aerodynamics, in: Lecture Series 2000-08, Verification
and Validation of Computational Fluid Dynamics, 5–8 June 2000, Von Kármán Institute for Fluid Dynamics, Rhode-Saint-Genèse, Belgium.
[17] E. Rosencher, Optical diagnostics of flows, Aerosp. Lab. J. 1 (December 2009), http://dx.doi.org/10.12762/2009.AL01.

JID:CRAS2B AID:3504 /SSU [m3G; v1.219; Prn:13/07/2017; 8:28] P.14 (1-14)
14 B. Chanetz / C. R. Mecanique ••• (••••)•••–•••
[18] S. Andrieux, Testing in aerospace research, Aerosp. Lab. J. 12 (December 2016), http://dx.doi.org/10.12762/2016.AL12.
[19] J.-M. Weber, Un demi-siècle d’aéronautique en France, Études et recherche, Comité pour l’histoire de l’aéronautique, Theatrum Belli, Bibliothèque
Défense et Sécurité, 2008.
[20] D.B. Baals, W.R. Corliss, Wind Tunnels of NASA, Scientific and Technical Information Branch, National Aeronautics and Space Administration, 1981.